The data landscape (continued)

advertisement
Data Reference
the data landscape
(continued)
Surveys
• Census
• American community
survey
• Current population
survey
• General social survey
• California poll
• Roper organization
• American National
Election Survey
Administrative
records
•
•
•
•
•
•
•
•
•
Vital Statistics
Health administration
Crime reports
Court reports
Education reports
SEC,FCC, FTC, etc.
Legislative bodies
Voting records
Official economic
statistics
Data
Statistics
Data
AGGREGATE
DATA
Statistics
http://www.bls.gov/webapps/legacy/cpsatab7.htm
8.1
7.9
8.5
Rate of
unemployment
20-24 yr-old
Aggregated
from data on
individuals
Data over time
http://www.bls.gov/webapps/legacy/cpsatab7.htm
Reformatted
for use
by
statistical
software
Rate of
unemployment
20-24 yr-old
Aggregated
from data on
individuals
Data over time
Typical Characteristics of
Aggregate Data




Contrast with microdata (lowest level of
observation, e.g., person, firm, event)
Summarize microdata over time or space.
Allow comparisons between geographic areas
or across time
File formatted for easy use by statistical
software.
Some Typical Aggregate Data



Census data ("summary files")
Economic data (time series)
Business data (company or industry
aggregates)
The real world
Data
(microdata)
Aggregate Data
Statistical tables
Scholarly literature
Popular press
Continuum of access

Microdata
–
–






Public use microdata
Confidential microdata
Aggregate data
Databases
eTables (digital, tabular, on the web or on CD, etc.)
ePublications (on the web or CD. Reports, statistical
abstracts, etc.)
Scholarly publications
Popular press
Continuum of access
Confidential microdata
Public use microdata
Aggregate data
databases
Scholarly literature
popular press
eTables
ePublications
Continuum of access
Confidential microdata
Public use microdata
data
Aggregate data
databases
Scholarly literature
popular press
(CONTENT)
eTables
ePublications
statistics
Continuum of access
Confidential microdata
Public use microdata
Aggregate data
databases
Scholarly literature
popular press
eTables
ePublications
data
(CONTENT)
statistics
less
(INDEXING)
more
Continuum of access
Confidential microdata
Public use microdata
Aggregate data
databases
Scholarly literature
popular press
eTables
ePublications
data
(CONTENT)
statistics
less
(INDEXING)
more
expensive
(COSTS)
free
Continuum of access
Confidential microdata
Public use microdata
Aggregate data
Scholarly literature
databases
popular press
eTables
ePublications
data
(CONTENT)
statistics
less
(INDEXING)
more
expensive
(COSTS)
free
restricted
(ACCESS)
open
Access Points
•News.google.com
•Lexis/Nexis
•Newsbank
•Factiva
•Proquest
•Etc…
Popular Press
Access Points
•Scholar.google.com
•Discipline indexes
(Population Index,
EconLit, PAIS,
Sociofile, etc.)
•Collection indexes
(Muse, World Development
Sources, Criminology, etc.)
Scholarly Literature
Access Points
•Statistical Universe
•Statistical compendia
and abstracts
•Bibliographies of
statistical publications
Statistics
Access Points
Aggregate data
Aggregate data distributors
•Fedstats: fedstats.gov
•American Factfinder: factfinder.census.gov
•International Financial Statistics: imfstatistics.org
•Rand California: ca.rand.org
•United Nations Common Database: unstats.un.org/unsd/cdb
•Bureau of Labor Statistics: stats.bls.gov
Access Points
Microdata
Data Archives and Data Libraries
•ICPSR: www.icpsr.umich.edu
•UK Data Archive: www.data-archive.ac.uk
•Council of European Social Science Data Archives: www.nsd.uib.no/Cessda
•History Data Service: hds.essex.ac.uk
•American National Election Studies: www.umich.edu/~nes
•Association of Religion Data Archives: thearda.com
•NARA: www.archives.gov/research/electronic-records
•More: odwin.ucsd.edu/idata
Tip: use the tools you know to find
statistics that will lead you to data





Bibliographic indexes (OPACs for books and
data, journal indexes, news indexes, statistical
indexes)
Web search engines
"Agency" approach (who might have collected
the data?)
Known compilations of statistics
Known data sources, archives, libraries
Tip: work "backwards" from
statistics to the data source




Footnotes
Bibliography
Captions under tables and graphs
text
http://www.census.gov/prod/www/abs/popula.html#pop
Tip: there is no one-and-only right
way to begin a search for data



Ask 3 data librarians where to start a datasearch and you'll probably get 3 different
answers.
Each reference provider brings his or her own
skills and experiences, knowledge and
strategies.
Corollary: There may be many right end points
of a search.
Question One

"I'm looking for data about trends in teenage
smoking."
Question Two

"Have there been more deaths due to hot
weather recently?"
What have we learned






Data come from surveys and administrative records
Statistics come from data
Aggregate data are summaries of microdata formatted
for use by statistical software
The "continuum of data access" provides many entry
points for searching for data.
The "best" search starts with tools that are familiar and
may lead to statistics first.
In 5 minutes or less we can get users focused and
started on their exploration for data.
Data Reference
Next time:
The data reference interview
And…
Cool tools and strategies
Download